{ "name": "Drevanto Beta 2 verification summary", "publisher": "Drevanto", "attribution": "Drevanto internal certification and verification. Not a third-party audit. The construction-reference campaign used independently sourced authoritative references; Drevanto performed the evaluation and scoring.", "lastVerified": "2026-08-29", "sanitization": "Public-safe aggregate summary only. Intentionally omits source-manifest hashes, source identifiers, local paths, raw fixtures, formulas, internal implementation details, private evidence filenames, device serials, tester identities, credentials, and customer data.", "gates": [ { "id": "instant-estimate-benchmark", "name": "Certified Instant Estimate benchmark", "date": "2026-08-21", "status": "PASS", "results": { "certifiedGenerations": 520, "supportedAssemblyVersions": 26, "combinedMedianSeconds": 0.028, "generationFailures": 0, "generativeAiProviderCalls": 0, "combinedP95Milliseconds": 48.008, "combinedP95SecondsRounded": 0.048 }, "limitations": [ "Certified supported jobs only.", "Does not certify supplier-price accuracy or real-world job coverage.", "Android figures come from an emulator, not a fleet of physical devices.", "Historical campaign: these are dated 2026-08-21 results on that source state, not a re-run against the current Windows Build 17 or any later build.", "Android figures come from an API 36 emulator; no physical handset was timed in this campaign.", "Timing starts when Generate is pressed after clarification and ends when the editable estimate is visible. Application startup and the clarification questions themselves are excluded." ], "methodology": "", "timingBoundary": "Timing starts when Generate is pressed after clarification and ends when the editable estimate is visible. Application startup and the clarification questions themselves are excluded.", "unsupportedScope": "Unsupported or unusual scope is not forced through Instant \u2014 it is flagged and routed to Drevanto's Hybrid/assisted review path for contractor review. In the current beta that path is manual contractor review: AI-assisted drafting requires an external provider and no provider is enabled, so no generative-AI step runs." }, { "id": "construction-benchmark-lab-v3", "name": "Independent Construction Benchmark Lab v3", "date": "2026-08-28", "status": "PASS", "methodology": "Drevanto output compared against independently sourced authoritative construction references until meaningful coverage saturation was reached. Evaluation performed by Drevanto.", "results": { "acceptedReferences": 250, "acceptedSourceDocuments": 92, "newReferences": 139, "newSourceDocuments": 43, "previouslyCertifiedCorpusRetained": 111, "tierCSourcesUsedAsTruth": 0, "scopeRecallPercent": 100.0, "scopePrecisionPercent": 87.91, "quantityMedianAbsolutePercentDifference": 0.0, "quantityP90AbsolutePercentDifference": 30.0, "professionalBoundaryViolations": 0, "confirmedDefectsRemaining": 0, "governedRepairs": { "count": 1, "independentlyCorroborated": true, "scope": "Scope-language recognition only. No change to quantities, prices, professional boundaries, or protected baseline artifacts." } }, "limitations": [ "Quantitative comparisons only cover cases where authoritative references establish explicit normalized quantities.", "Legitimate construction methods can differ; a difference is not automatically an error.", "Coverage saturation does not mean every construction reference in existence was consulted." ] }, { "id": "guided-input-composed-projects", "name": "Guided numeric-input and composed-project certification", "date": "2026-08-29", "status": "PASS", "methodology": "Real Guided UI controls exercised character by character, then driven through composed multi-part projects end to end.", "results": { "acceptedGovernedEntries": [ 2, 24, 100, 24.5, 24.25 ], "partialNumericTextCreatedConfirmedFact": false, "explicitCommitRequired": true, "invalidInputsMutatedConfirmedFacts": false, "composedScenariosCompleted": [ "attached-garage addition", "existing-roof tie-in" ], "descriptionSuppliedInformationReAsked": false, "prematureGenerationReady": 0, "readinessMismatches": 0, "orphanBlockers": 0, "deadControls": 0, "missingInformation": 0, "generationSucceeded": true, "professionalAndFieldVerificationBoundariesPreserved": true }, "limitations": [ "Covers the certified Guided flows and scenarios listed, not every possible authoring sequence." ] }, { "id": "exhaustive-modeled-transition-closure", "name": "Exhaustive modeled transition closure", "date": "2026-08-29", "status": "PASS", "results": { "modeledStates": 20, "registeredOperations": 68, "testedModeledEdges": 983, "untestedCells": 0, "pairwiseObligations": 1197, "highRiskThreeWay": 42, "highRiskFourWay": 4, "productionProbeFailures": 0, "longSessionTransitions": 100000, "longSessionArchetypes": 18, "failureSignatures": 0, "stateCoverageSaturationReached": true, "duplicateGenerationAttemptsPrevented": 19800, "persistenceChecks": 10000, "generationChecks": 15000, "nextQuestionChecks": 5000, "pathologicalSlowdown": false, "runawayMemoryGrowth": false, "deadlock": false, "unboundedQueueGrowth": false, "externalProviderCalls": 0 }, "limitations": [ "Closes the modeled workflow; does not model sequences outside that model." ] }, { "id": "all-trade-big-box-final-boss", "name": "All-Trade Big-Box Final Boss", "date": "2026-08-29", "status": "PASS", "methodology": "Synthetic integration fixture: a fictional 120,000 square foot large-format retail project. No real company, plan, price, or confidential information.", "results": { "reproducibleCleanSemanticRuns": 2, "supportedScopesDetectedAndCoordinated": 20, "responsibilityRows": 20, "takeoffRows": 35, "missingInformation": 0, "generationCount": 1, "finalFailures": 0, "gatesPassed": [ "Reviewer", "Margin Guardian", "Job Ready", "ownership/duplication", "mutation", "save/resume", "cross-trade interface" ] }, "limitations": [ "Synthetic integration fixture. Not proof that every real project or price is correct." ] }, { "id": "final-release-candidate-validation", "name": "Final release-candidate validation", "date": "2026-08-29", "status": "PASS", "results": { "android": { "version": "1.2.0-beta.2+6", "exactSignedReleaseCandidateVerified": true, "packageVersionSigningArchiveSourceBinding": "PASS", "playUploadCertificateCompatibility": "PASS", "bootstrapPluginRegistration": "repaired and deterministically verified before retest", "physicalDevice": "Samsung SM-S938U (Android 16)", "physicalRetest": "PASS", "priorColdLaunchPluginRegistrationCrash": "confirmed resolved", "coverage": [ "repeated cold launch", "authentication and session persistence", "Dashboard", "Customers", "Estimates", "Smart Estimate Builder", "character-by-character multi-digit input", "composed attached-garage/addition Guided UX", "description consumption", "readiness and build gating", "successful generation", "save and reopen", "Jobs", "Schedule", "Reports", "Settings", "force-stop and relaunch", "crash monitoring", "cleanup of temporary test data" ] }, "windows": { "version": "1.2.0-beta.2+5", "exactPackagedInstallerVerified": true, "sourceBinding": "PASS", "integrityChecks": "PASS", "authenticodePublisher": "ARMM General llc", "coverage": [ "isolated installation", "cold launch", "authentication and session behavior", "main workspaces", "numeric input", "composed garage/addition flow", "generation", "save, reopen, and delete a temporary estimate", "close and relaunch", "crash monitoring", "rollback and uninstall safety" ] }, "releaseArtifacts": { "archiveIntegrity": "PASS", "forbiddenPackagedFilesFound": 0, "privateHashesLocation": "protected release evidence and download surfaces only" } }, "limitations": [ "Covers the certified beta candidates on the hardware tested, not all possible hardware, OS builds, or field conditions." ] } ], "overallLimitations": [ "Drevanto internal certification and verification, not a third-party audit or industry certification.", "Does not certify supplier-price accuracy, local code compliance, structural adequacy, or real-world job frequency and coverage.", "Output is editable; contractor pricing, review, and field verification remain in the contractor's control." ], "correctionNote": "Correction (September 7, 2026): earlier copy stated a flat \u201c95% completed within 0.048 seconds\u201d. The measured 95th percentile is 48.008 milliseconds; 0.048 seconds is a rounded display value. Download URLs are unchanged." }